ABSTRACT Making a correct diagnosis in a timely manner and ensuring that it is linked with the correct treatment in acute care represents an unresolved patient safety issue. If the diagnosis is not correct, the patient will likely not get the correct therapy, which can result in harm. Even when the diagnosis is correct, all too often the correct therapy is not delivered quickly, effectively or appropriately. To address this overall issue, we will utilize rigorous systems engineering and human factors methods to guide our approach. We will begin with problem analysis, then design and develop the intervention, iteratively improve it, and evaluate it, include evaluating our ?system-of-systems?. We will select a number of common, costly diagnoses or situations which have a high likelihood that either the diagnosis or treatment are not correct. We will then identify and assess several triggers which suggest a problem with either the diagnosis or therapy, such as failure to respond in a diagnosis-specific timeframe. From our current Patient Safety Learning Laboratory and other work, we have built a variety of technological approaches which we can use to interact with patients and providers. For example, we developed a safety dashboard, integrated with our electronic health record, which is routinely used as a checklist to ensure safety during delivery of care in our hospital. We will also ask patients (and their caregivers as appropriate) whether they are concerned that their diagnosis or treatment may not be correct, and share that with the care team. We will provide team-based training to create a culture of diagnostic safety on the clinical units. We will then measure whether or not this results in fewer diagnostic or therapeutic errors, both overall and by condition.